Iodide removal

The I⁻ removal efficiency with a variation of BPA dosages is shown in Figure 2.11. An increase in the cationic BPA dosage leads to greater I⁻ adsorption within 2 h of contact time.

At 2 g/L, all BPA can remove up to 50% of the 10 mg/L I⁻ in water. At 5 g/L, PA-R1, PA-R2, and PA-R4 tend to remove 92 – 95%, while PA-R3 can remove 80% of the same I⁻

concentration. At a 10 g/L dosage, all BPA could remove 10 mg/L I⁻ from the solution except PA-R3, which adsorbed about 95% of the initial concentration. Considering the limiting BPA concentration, in the kinetic experiment, a 10 g/L dosage was selected for PA-R3, and 5 g/L was used for the other BPA.

Figure 2.11 Dosage variation of cationic biopolyamide in removing 10 mg/L I⁻ for 120 min under room temperature

The I⁻ adsorption performance with a 3-hour contact period of the BPA is presented in Figure 2.12a. The equilibrium time of contact was determined to be approximately 120 min.

35

PA-R1, PA-R2, and PA-R4 showed a similar I⁻ removal pattern, with nearly 100% at equilibrium. PA-R3 I⁻ adsorption efficiency reached the equilibrium within the same duration but with a lower efficiency (almost 80%). At the optimum time, 0.8 mg/g adsorption capacity was observed for PA-R3, while the other types of cation polyamides removed up to 2 mg/g (Figure 2.12b). The lower efficiency of PA-R3, compared to that of other BPA, may be due to its lower Mw. The molecular weight may affect the physical adsorption of I⁻, such that at a low concentration, a higher molecular weight polymer allows for more coiling of polymer chains, leading to entrapment of more I⁻.⁸⁴

Figure 2.12 I⁻ removal of cationic BPA over time using optimum dosage under room temperature. (a) percent removal efficiency calculated by Equation 1, (b) adsorption capacity

The numerical values for linear fits of pseudo-order-reaction rates are tabulated in Table 2.4. The actual fitting of the first- and second-order can be found in Figure 2.13a and Figure 2.13b, respectively. It was observed that the logarithmic plot for the I⁻ removal of PA-R3 was almost linear (R² = 0.994) with time, and therefore, followed the pseudo-first-order reaction.

The linear first-order kinetics depicted the anion-exchange reaction that was proportional to only the amount of I⁻, and not the amount of adsorbent.⁸⁵ In contrast, R1, R2, and PA-R4 fitted both order reactions, inferring that their adsorption mechanisms are influenced by the amount of both the solute and the polymer.⁸⁵ A combination of chemisorption and

36

physisorption for the I⁻adsorption by the bio-based cationic polymer could be deduced from the results of the pseudo reaction order fitting.

Table 2.4 Derived parameters of the linear fitting pseudo-order reactions and isotherms for the I⁻ removal by cationic BPA

catio -nic BPA

pseudo-first-order reaction^a pseudo-second-order reaction^b Langmuir isotherm^c Freundlich isotherm^d

k1

(L/min) Qe

(mg/g) R² k2

(g/mg/min) Qe

(mg/g) R² KL

(L/g) qmax

(mg/g) R² KF

(L/g) n R²

PA-R1 0.046 1.64 0.952 7.60 × 10^-3 2.57 0.971 n/a n/a n/a n/a n/a n/a

PA-R2 0.033 1.42 0.987 0.051 2.50 0.988 n/a n/a n/a n/a n/a n/a

PA-R3 0.012 0.92 0.994 7.36 × 10^-6 30.6 0.565 n/a n/a n/a n/a n/a n/a

PA-R4 0.046 1.57 0.958 8.10 × 10^-3 2.51 0.972 0.078 32.3 0.998 0.032 0.2 0.977 aparameters calculated by Equation 2, ^bparameters calculated by Equation 3, ^cparameters calculated by Equation 5, ^dparameters calculated by Equation 6

n/a – not available

Figure 2.13 Linear fits of pseudo-order-reaction for the I⁻ removal by cationic biopolyamide.

(a) first-order, (b) second-order

37

PA-R4 was further tested for the batch experiment due to its performance, and the maximum q obtained from the experiment is about 30 mg/g (Figure 2.14a). The fit of linear isotherm is depicted in Figure 2.14b, where the distribution coefficient (KD) is calculated to be 7.58 L/g. The value was similar to that of various types of polystyrene-based resin.⁷⁰ The fittings of Langmuir and Freundlich isotherms are shown in Figure 2.14c and Figure 2.14d, respectively. The qmax value of 32.3 mg/g yielded by the Langmuir model (Table 2.4) is found to be similar to that of the mentioned experimental maximum q (Figure 2.14a). In comparison with other bio-based adsorbents from the literature, which were observed to remove I⁻ in similar experimental conditions, the performance of the presented cationic BPA is deemed acceptable. The reported values of qmax for copper-modified activated carbon and TiO2-Fe2O3 -PVA alginate beads were 1.91 and 20 mg/g, respectively.^65,86 PA-R4 was best represented by the Langmuir isotherm with an R² of 0.998 (Table 2.4 and Figure 2.14c ), which inferred that there was a conformity of ion-exchange reactions in all sites of the cationic polyamide.⁸⁷ With an R² of 0.977 (Table 2.4 and Figure 2.14d), the Freundlich model too, can used to predict the behavior of PA-R4. The model yielded 0.2 for n, which means the free adsorption energy is constant throughout different I⁻ concentrations.⁸⁷

The PA-R4 results for the 10 mg/g I⁻ adsorption efficiency at various temperatures are shown in Figure 2.15a. At 20 °C, DOWEX^TM 550 was found to be superior for I⁻ adsorption, as it has 100% I⁻ removal capacity, while PA-R4 I⁻ adsorption was up to 98%. An increase in the temperature led to a significant loss in the I⁻ adsorption efficiency for the conventional resin. In contrast, the BPA maintained their anion-exchange efficiency throughout the measured temperature range, and the performance began to fall after 80 °C. From 40 °C, except for PA-R3, the I⁻ removal efficiency of BPA started to surpass that of DOWEX^TM 550. The lower efficiency of the commercial-grade resin was may be attributed to the low thermal stability of the polystyrene resin (Table 1.1). Moreover, the water absorption characteristics of

38

the BPA may increase the OH⁻ (H2O)x clusters in the complex, subsequently preventing the degradation of quaternary ammonium to further facilitate the uniform adsorption of the dissolved impurities at a higher temperature.^72,88 This phenomenon was less observed in a polystyrene-based material like Dowex^TM 550, as reported during water immersion tests.⁸⁹

Figure 2.14 PA-R4 isotherm fits. (a) adsorption capacity over final I⁻ concentration, (b) linear isotherm, (c) Langmuir isotherm, (d) Freundlich isotherm

All the cationic polymers exhibited the same trend of I⁻ removal efficiency in various pH condition (Figure 2.15b). The highest removal was observed at pH 2 and remained stable until about pH 7, where the decrease in I⁻ removal occurred. The acidic condition promoted the sorption of I⁻ by positive charges in the sorbents, while under high pH condition, the presence of OH⁻ would create a competition for iodine anions.⁹⁰ While the performance of PA-R1 and PA-R2 was drastically affected from the pH increase, especially from pH 10 onwards,

39

PA-R4 and the commercial-grade resin were perceived to be stable with more than 80% I⁻

removal at pH 12.

The selectivity of PA-R4 for I⁻ was briefly accessed (Figure 2.15c). In the presence of other anions, PA-R4 could maintained the I⁻ removal efficiency above 70%. The adsorption efficiency was most affected by SO42⁻. The efficiency of almost 100% was observed in the tap water, while about 50% could be achieved in the seawater where numerous competitive anions existed in high concentration.

Figure 2.15d depicted the reusability of the PA-R4 for 5 cycles. The I⁻ adsorption efficiency was maintained above 90% in the second cycle. The fifth cycle of regeneration and I⁻ adsorption yielded an efficiency of less than 50%.

All in all, our research has produced a series of cationic BPA with high thermal properties and ability to remove I⁻ from the water with a performance comparable to that of conventional polystyrene resin and metal-inclusive adsorbents. The reusable PA-R4 is not only considered to be fully bio-based, but its anion-exchange processes in the water were observed to be stable throughout different temperature and pH. It also has a fair selectivity toward I⁻ and is reusable after KOH regeneration.

40

Figure 2.15 I⁻ removal performance of BPA in different background conditions and regeneration cycles. (a) temperature variation [20 – 90 °C], (b) pH variation [2 – 12], (c) selectivity of PA-R4 for removal of I⁻, (d) performance of regenerated PA-R4 for 5 cycles. The experiment was performed with 5 g/L BPA, 10 mg/L I⁻, and 120 min of contact time. The regeneration by 1 M KOH took 6 h. The conventional cationic resin, DOWEX^TM 550, was used for comparison in the temperature and pH variation experiment

41